Nielsen is still the top dog in the world of television-audience measurement. But in some ways, that audience is running away. Nielsen — which, as a company, turned 90 years old this year — has a virtual monopoly on the national and local TV ratings that are the currency used to determine how some $63.8 billion is spent yearly on TV advertising in the United States alone.
Extrapolating data collected overnight from more than 22,000 homes across the country, with meters connected to each television recording what each person in the home views, Nielsen churns out reports that tell networks and advertisers in detail how many people, by age, gender and ethnicity, were watching any of 132 currently reported broadcast and cable channels any hour of the day. Over the years, other companies, including Arbitron (which Nielsen has a pending deal to acquire), U.K.-based AGB Group (which Nielsen has partnered with since 2004) and startup Statistical Research Inc. have tried to mount serious challenges to Nielsen’s national ratings business — but none have succeeded.
Despite an increasingly fragmented audience, and occasional loud complaints of undercounted audiences from such networks as Nickelodeon and Univision, Nielsen’s ratings continue to reign supreme.
Today a wave of newer, more nimble companies are swooping in to try to fill in the gaps left by Nielsen’s measurement, including Rentrak and TiVo, which have created measurement services using set-top data. And while the set-top clicks don’t provide the demographic information Nielsen gleans from viewing in its homes, they do provide a clear picture of where homes are and what their occupants are watching. (Nielsen augments local ratings already with set-top return path data from DirecTV and Charter Communications.)
Advertisers’ desire to go beyond TV ratings to launch crossplatform campaigns reaching online, video-on-demand and linear television have forced them into creating hybrid “Frankenmetrics,” combining measurement data from a variety of sources, network officials have complained. (Nielsen said it is addressing those gaps with clients, including testing a system called Digital Program Ratings that includes online views of network content.)
As more viewers chat and tweet about shows — and even during them — several other competitors have materialized to filter and analyze this data as well. The effect of social media on TV ratings represents an arena that may someday create a new currency for buying and selling ads, some believe. The logic of Nielsen’s recent pre-emptive partnership with Twitter was backed by its own research that showed tweets could help lift live-TV ratings, especially for competitive reality shows.
Despite complaints about the ratings status quo, providers such as Comcast have said Nielsen is making progress in expanding into better measurement of VOD views and into the as-yet unmeasured realm of TV viewed on such mobile devices as tablets and smartphones.
Measurement issues haven’t been holding back adoption of TV Everywhere services, Comcast senior vice president of digital and emerging platforms Matt Strauss said earlier this summer at an industry event. “From a product standpoint, the metrics are getting there,” he said. “It’s not slowing down our ability to move quickly.”
On the video-on-demand front, Comcast and Nielsen have been testing different digital ad-insertion techniques to try to add to the ad-sales haul that VOD viewing is currently contributing as part of the C3 views in the three days after a program airs.
While Nielsen feels the household panel data is still the gold standard, “we know that there’s no way that we can build these panels large enough — as we project out in the future — to be able to collect the granularity of information that we want to work with,” Brian Fuhrer, senior vice president of product research at Nielsen, told Multichannel News. “The panel will always have value for certain views of the data. But we have to be really smart about how we augment the panels or how we use the panels for calibration tools for the broader data collection we have out there. The technology is really providing new ways to collect the data.”
Measuring mobile viewing — on iPads, other tablet devices or smartphones — is Nielsen’s most pressing need, based on what it hears from its clients, Fuhrer said. Those are the big views that are currently getting away.
“As far as an overall priority, that’s probably the biggest one on the radar screen,” Fuhrer said. “We’re calling it our big bang approach, of trying to really expand measurement. We feel that’s very, very critical. We want to make sure that we’re doing everything we can to not only support the existing business models, but also reflect the total landscape of how consumers are accessing video.”
Surveys and anecdotal indicators abound as to the growth of viewing on iPads and other mobile devices — viewing that Nielsen, despite its 22,000-home panel with “people meters” hooked to every TV and its 200,000-user panel of online video watchers, can’t yet track.
A Leichtman Research Group survey in May found 27% of adults in the U.S. watch video on devices other than a TV set every day, and 53% do so on a weekly basis. That’s up from 14% daily and 37% weekly two years ago.
When that viewing is done on an iPad or an eReader, 89% of it is usually done at home, LRG president and principal analyst Bruce Leichtman said.
The percentages of households watching video on non-video devices is growing, but “it still represents a very small amount of all video viewing,” Leichtman said. “That said, the vast majority of this viewing is being done in the home (essentially using iPads/tablets, and smartphones as additional outlets), so for all involved, I think that the answer would be, ‘The sooner the better.’ ”
More TV viewers are also watching content delivered via connected devices such as Microsoft’s Xbox, Apple TV and Roku set-tops. Nielsen this fall will add views from broadband-connected TV sets to address that development, though that is expected to have little impact on overall ratings, because less than 1% of TV homes meet the new definition.
Based on recent briefings Nielsen executives provided at the company’s Global Technology and Information Center, a 611,000-square-foot complex near Tampa, Fla., where the overnight data is processed and analyzed and field representatives that manage the sample homes are trained (see sidebar), here is a look at what Nielsen is doing on technology fronts to keep pace with current and anticipated needs in TV measurement.
TABLET VIEWING: TV viewing on a computer is being measured in some Nielsen households today, using installed software that captures watermarks and audio signatures in the TV signal and sends them to Nielsen over the Internet. For mobile devices such as tablets or smartphones, rather than installing software, Nielsen’s preferred approach is to insert watermarks into ID3 tags, which are containers for metadata for digital content such as MP3s. Apple’s iOS and Google’s Android devices already support ID3 tags. The Nielsen ID3 tags would need to be integrated into the video content and a corresponding reader into the video player apps. Nielsen has done some testing of the technology, including a reported trial this past spring involving four CBS stations and software firm Syncback. Nielsen hopes to have some proprietary data to share with clients this fall and has said the goal is to have viewing data that fits the C3 criteria for commercial-minute ratings by fall 2014. Nielsen also expects to augment mobile viewing data from household panels with other sources, Fuhrer said, rather than having to ensure that enough of the 22,000 homes in the nationwide panel have tablets or smartphones.
VOD MEASUREMENT: C3 ratings currently include TV viewing that takes place over three days after a program airs, including digital video recorder playbacks and video-on- demand orders. Nielsen said it is now able to isolate how much of that viewing is done on VOD if networks separately encode on-demand telecasts. Eight networks, mostly in the Comcast/NBCUniversal camp (Nielsen won’t spell them out), began participating in that exercise in separate encoding and measurement in June. “We anticipate some really good insights there,” Kevin Rini, director in media product leadership, said in a May interview. Early looks at the data showed some programs were seeing more than 10-15% of their views on VOD, while a more typical result was about 3-5%, Fuhrer said. Nielsen and Comcast want to expand VOD measurement to include library episodes of a show, if they have had current ads digitally inserted and are seen in the same three-day period as the most recent telecast. The two companies are testing that approach in Nielsen test homes in two (unnamed) Comcast markets but have not disclosed any results.
NEW METERS: The next generation of Nielsen-household meters will be smaller and won’t need to be wired directly into a television or DVR or other viewing device. They measure audio “watermarks” or tones inserted into the TV stream and are encoded with date, time and program information. They will be easier to install and to interact with than current “active/passive” or A/P meters. Called “GTAM,” for global television audio metering, the new system also is intended to standardize Nielsen meters around the world on the latest technology, including watermarks, which are a newer version of the audio signatures Nielsen also measures. An upgraded version of the current Nielsen people meter, called a scrolling text people meter, uses text displays to replace the blinking lights now used to prompt viewers to activate the device. A panel of 300 homes is testing the new GTAM meters, and toward the end of the year a side-by-side analysis will be done to compare GTAM and A/P viewing data. The new meters are slated to be deployed in 2014. People meters, used in the top 25 TV markets, record demographic data about viewing by different members of the household. In the next tier of markets, Nos. 26-56, Nielsen homes use set meters (that record what is being viewed but not demographic breakdowns) plus diaries kept by viewers. In markets 57-210, viewer diaries are used solely, no meters.
Ratings-measurement giant Nielsen is making strides to stay with the TV audience as they shift viewing to emerging platforms.
GETTING HOMES TO SAY ‘YES’ TO NIELSEN
OLDSMAR, Fla. — The care and feeding of 22,000 households (with 55,000 people) in Nielsen’s measurement samples across the country require hundreds of people who can get a “yes” answer — if not the first time, then eventually.
As to feeding, Nielsen membership representatives who knock on doors that have been pre-identified as prime targets don’t come empty handed. Oft en, they bring food.
“Cracker Barrel tends to be kind of popular,” Fred Rogers, vice president of U.S. media field operations for Nielsen, said. Given regional variations, the $10-$15 warmup gift might be a Starbucks basket in Seattle or cheesecake in “New Yawk,” as Rogers put it.
Nielsen selects target homes, or “basics,” on the basis of geography, Rogers said. “Every home in the United States has an equal probability of being chosen.” The number of homes in a market will vary based on population: Tampa, Fla., has about 600 homes in the panel, while New York and Los Angeles have 1,000.
Actually, Nielsen wants to have more than that number online to guard against inevitable “sample impacts,” ranging from overnight power outages to evictions. So, the idea is to have 610 active homes in the sample in Tampa and 1,020 in L.A.
Before the rep shows up bearing Cracker Barrel and looking for a “yes,” the household would have been mailed a questionnaire accompanied by 10 one-dollar bills. About half of those questionnaires are filled out and mailed back, Rogers said, at which time Nielsen mails back a $20 fulfillment fee.
Nielsen gets about six out of 10 “basic” homes to agree to participate. It continues to follow up with the four homes that say no, hoping for a change of heart or occupant. Until the basics are signed up, Nielsen recruits “surrogates” to fill in.
For homes that agree to take part, there is about a six-hour installation process, at least in the top 25 designated market areas, or DMAs, in which people meters are attached to televisions and other devices, such as Xboxes, used to view TV. Membership representatives collect about 500 pieces of data on each household; people meters prompt viewers to log in and identify who is watching. Data is collected and sent overnight, mostly over cellular networks. Reps spend a portion of that six-hour install coaching on people meter use.
Nielsen homes are paid a small sum (about $50) when they sign on, then a similar stipend every six months after that. The money typically isn’t a big motivator, Rogers said. A bigger incentive is contributing to an accurate view of what Americans watch on TV.
Membership reps, who receive about three weeks of training in Florida, are prized for their ability to quickly be seen as trustworthy and communicate with people from all walks of life. “At the end of the day, it’s about putting someone in front of you that you will say yes to having equipment put in your home,” Rogers said.
Nielsen homes age out after two years. Field reps, who maintain Nielsen gear in the home, can’t take vacations around the December holidays, when some 5,000- 7,000 new TVs, Xboxes must be wired into the system.
— Kent Gibbons